Artificial intelligence for detection of microsatellite instability in colorectal cancer-a multicentric analysis of a pre-screening tool for clinical application.

BACKGROUND: Microsatellite instability (MSI)/mismatch repair deficiency (dMMR) is a key genetic feature which should be tested in every patient with colorectal cancer (CRC) according to medical guidelines. Artificial intelligence (AI) methods can detect MSI/dMMR directly in routine pathology slides, but the test performance has not been systematically investigated with predefined test thresholds. METHOD: We trained and validated AI-based MSI/dMMR detectors and evaluated predefined performance metrics using nine patient cohorts of 8343 patients across different countries and ethnicities. RESULTS: Classifiers achieved clinical-grade performance, yielding an area under the receiver operating curve (AUROC) of up to 0.96 without using any manual annotations. Subsequently, we show that the AI system can be applied as a rule-out test: by using cohort-specific thresholds, on average 52.73% of tumors in each surgical cohort [total number of MSI/dMMR = 1020, microsatellite stable (MSS)/ proficient mismatch repair (pMMR) = 7323 patients] could be identified as MSS/pMMR with a fixed sensitivity at 95%. In an additional cohort of N = 1530 (MSI/dMMR = 211, MSS/pMMR = 1319) endoscopy biopsy samples, the system achieved an AUROC of 0.89, and the cohort-specific threshold ruled out 44.12% of tumors with a fixed sensitivity at 95%. As a more robust alternative to cohort-specific thresholds, we showed that with a fixed threshold of 0.25 for all the cohorts, we can rule-out 25.51% in surgical specimens and 6.10% in biopsies. INTERPRETATION: When applied in a clinical setting, this means that the AI system can rule out MSI/dMMR in a quarter (with global thresholds) or half of all CRC patients (with local fine-tuning), thereby reducing cost and turnaround time for molecular profiling.

The founder mutation MSH2*1906G-->C is an important cause of hereditary nonpolyposis colorectal cancer in the Ashkenazi Jewish population.

Hereditary nonpolyposis colorectal cancer (HNPCC) is caused by mutations in the mismatch-repair genes. We report here the identification and characterization of a founder mutation in MSH2 in the Ashkenazi Jewish population. We identified a nucleotide substitution, MSH2*1906G-->C, which results in a substitution of proline for alanine at codon 636 in the MSH2 protein. This allele was identified in 15 unrelated Ashkenazi Jewish families with HNPCC, most of which meet the Amsterdam criteria. Genotype analysis of 18 polymorphic loci within and flanking MSH2 suggested a single origin for the mutation. All colorectal cancers tested showed microsatellite instability and absence of MSH2 protein, by immunohistochemical analysis. In an analysis of a population-based incident series of 686 Ashkenazi Jews from Israel who have colorectal cancer, we identified 3 (0.44%) mutation carriers. Persons with a family history of colorectal or endometrial cancer were more likely to carry the mutation than were those without such a family history (P=.042), and those with colorectal cancer who carried the mutation were, on average, younger than affected individuals who did not carry it (P=.033). The mutation was not detected in either 566 unaffected Ashkenazi Jews from Israel or 1,022 control individuals from New York. In hospital-based series, the 1906C allele was identified in 5/463 Ashkenazi Jews with colorectal cancer, in 2/197 with endometrial cancer, and in 0/83 with ovarian cancer. When families identified by family history and in case series are included, 25 apparently unrelated Ashkenazi Jewish families have been found to harbor this mutation. Although this pathogenic mutation is not frequent in the Ashkenazi Jewish population (accounting for 2%-3% of colorectal cancer in those whose age at diagnosis is <60 years), it is highly penetrant and accounts for approximately one-third of HNPCC in Ashkenazi Jewish families that fulfill the Amsterdam criteria.